Behavioral and Structural Model Based Approaches to Discrete Diagnosis

Abstract: The basic motivation for this thesis is the fact that things go wrong. With the growing complexity of todays engineering systems, the need has arisen for systematic approaches to failure diagnosis, i.e., fault detection and isolation.In the first part of this thesis an approach for modeling and diagnosis of systems that fall in the area of discrete event dynamic systems is proposed. The approach is applicable to systems that at some level of abstraction have an interesting discrete event dynamics that can display faulty behavior. The systems suitable for this approach typically consist of several interacting components where abrupt, butnon-catastrophic, faults can occur in the components.We use a relational framework for discrete event dynamic systems focusing on a conceptually simple representation of the relationship between inputs, outputs and states of a discrete event system. Faults and faulty behavior are modeled locally using the state variables, and the diagnosis problem basically is to infer the possible states of the system using the system model and observations of the real system, i.e., an observer problem. Detectability and isolatability properties are defined and algorithms for analysis are proposed. The transitions necessary and sufficient for detection can automatically be computed from the system model under certain conditions. We also show how to compute the nest possible fault partition.The second part of this thesis addresses the problem of fault propagation between software modules in a large-scale control system with object oriented architecture. There exists a conflict between object-oriented design goals such as encapsulation and modularity, and the possibility to suppress propagating error conditions. When an object detects an error condition, it is not desirable to perform the extensive querying of other objects that would be necessary to decide how close to the real fault the object is and hence whether it should report to the user.The fault propagation manifests itself as many irrelevant error messages and hence causes problems for system operators and service personnel trying to quickly isolate the real fault. A system developer with insight in the internal system design, can, of course, often easily interpret the multitude of error messages from a fault scenario and isolate the root cause. The key observation is that it can often be done using mental high-level models of the system and the mechanics of the fault propagation. We have made an effort to automate this procedure, and propose a fault isolation scheme as an extra layer between the operator and the core control system. In the fault isolation layer, post-processing of the fault information from the system is performed, to achieve clear and concise fault information to the operator without violating encapsulation and modularity.A high-level and informal explanation model for the fault propagation is presented and a taxonomy for error conditions in an object oriented system is proposed. We present algorithms and methods that use the explanation model and the error condition taxonomy together with a structural system model to form a cause-effect relation on the error messages, that can be used to find the most significant error message(s) in a fault scenario.